Airborne Imaging LIDAR for Biomass Estimation

There is a need for accurate and cost effective airborne techniques to estimate fish biomass in the open ocean. Arete proposes to develop in Phase I optimal detection algorithms using existing imaging LIDAR data collected with the Fisheye(R) sensor. These algorithms will be developed to allow real-time detection of fish and fish schools, discrimination between fish and ambient clutter, and discrimination between various fish species. This effort includes review and selection of VHS data, digitization of the data, development of reliable and efficient algorithms using statistical hypothesis testing, and scaling the algorithms to commercially available processors suitable for real-time operation. In Phase II, Arete would build and test the real-time processing system for use with the Fisheye(R) sensor. Phase III would lead to commercial applications in the fishing industry and in other industries that require real-time detection and recognition. The proposed recognition system, when coupled to an imaging system such as Fisheye(R), will detect and identify schools of fish from the surface down to 120 ft. Since Fisheye(R) is an active, gated sensor, depth-localization and day/night operation are possible. The operator will be alerted visually and audibly when a school of fish or individual fish has been found, allowing for the review of detection data via user-selectable display options, which will mitigate against clutter and false alarms. The proposed system can be used on existing airborne platforms (helicopter or small fixed-wing aircraft) and is adaptable to any imaging system that provides digital or analog data. Thus, the system will feature broad marketability and will be readily realized for use in commercial fishing fleets and scientific endeavors.